Lithium Ion Batteries for lights

[Burner]

Is there any reason not to use lithium ion for dive lights. Ignore the difficulties charging and expense. I have enough of them to make a 11v 10Ahr part that is 8" long in 2" pvc pipe.

Another thread mentioned it may explode when it comes in contact with salt water. It wouldn't be that bad to totally seal the battery so that nothing could ever get in. This might have other negative effects.

 

[cool_hardware52]

Lithium Ion batteries have been used in submerged apps with success.


I know you asked we ignore the expense and charging issues......but they are significant.

Ok, Ok lets ignore the $$$.

The real concern with Lithium Ion batts is limiting the max charge voltage and min discharge voltage, i.e. you need both a high and low limit.

Depending on the exact cell the high is 4.1-4.2 vdc per cell and the low is typicall 3.0.


As long as your pack has auto limits for upper and lower cutoff, and some form of over current protection (fuse or the like) I see no reason that lithium can't be used.

Many lithium cells have internal fuses to prevent catastrophic failure due to dead (crowbar) shorts. I'd guess a dead short is about the only way a flooded canister would lead to explosion.

If you use internally fused cells your pretty safe. Physically crushing the cell could still cause an internal short, but if something crushes your canister and the batts inside battery failure is likely not at the top of your list of troubles.




regards,


ch

 

[Charlie99]

One tidbit beyond CoolHardware's info: since the LiIon battery voltage changes signficantly during the discharge curve, it would be best to use a simple DC/DC converter to provide a constant voltage to the light.

 

[cool_hardware52]

Good point Charlie,

My day job, for last 18+ years has been designing and building High End bike lights.

Our lithium Ion system includes just such a DC/DC boost type voltage regulator.

Two main problems with Lith Ion, the voltage curve you mentioned and the available ballast voltages for HID.

Currently HID ballasts are available in 10.8 (9 cell Ni-Cad or Nimh)
12 (10 cell) 13.2 (11 cell), or 14.4 (12 cell)

Only the 14.4 matches well with the nominal voltage of Lith Ion cells (3.6 nom x 4 =14.4)

Our bike light uses a 13.2 volt ballast and a ~13 volt constant voltage boost regulator. Staying on the conservative side of the ballast input voltage reduces heat, increases life of both the ballast and lamp.

I will say that a DC-DC boost regulator with High efficiency, low Q-current and some means of indicating end of charge of the battery was not a trivial problem, made some what more complicated by our requirement that it handle at least 5 amps.


Regards,



ch

 

[Burner]

Does the ballast actually need a constant voltage to it? I thought it regulated the voltage to the light. I know that the SLA batteries I'm currently using have a discharge curve, I haven't looked to see how similiar it is to mine, but I don't think it will end up mattering too much.
I already have the batteries and the controller circuit and charger and stuff so cost doesn't really matter.
Thanks for the help.

 

[Charlie99]

The voltage of the Li Ion cells change much more over the useful discharge curve than does the voltage of NiCad, SLA, or NiMH batteries.

You didn't mention what type of bulb. I assumed basic incandescent halogen type. The light output and life of the bulb vary dramatically with voltage, particularly when you start cranking up the voltage and overdriving the bulb to get lots of light and a high color temperature. In this case, it is important to be able to set the voltage to the bulb independently from the incoming battery voltage.

There are some HID ballasts that can accept a wide input range, but the common 2 transistor "Royer" transformer driver circuits expect a fairly constant voltage.

 

[cool_hardware52]

Welsh Allen does publish a max and min voltage for all their HID ballast. My experience is that the max voltage is very important, do not exceed. Many of the ballast will continue to work below their min voltage, with a concurrent reduction in lamp output.


The rough rule of thumb for incandescent lamps is:

A 20% increase in voltage above the design voltage of the lamp (for ex a "10 vdc lamp" operated at 112 vdc) will result in a doubling of light output, but at the cost of reducing the lamp life to ~5-7% of design life. In other words a "100 hour bulb" run 20% above design voltage can be expected to fail in 5-7 hours.

A 20% reduction in voltage, (for ex a 10 vdc bulb run at 8 volts) will result in a 50% reduction in light output, and lamp life will become very extended, ~1000%.

What occurs in incandscents is a shift in output spectra. Lower voltages yeild more infared, long wave heat, higher voltages yeild more visible light.

With respect to batteries is "Area under the curve" If one plots voltage vs time and uses a max charge voltage and min discharge voltage as end points on the curve the area under the curve is the useable charge.

If the "top line" of this area is relatively flat and straight it's pretty easy to select and power either a incandscent lamp, or an unregulated ballast.

Ni-Cad, Nimh and SLA do have pretty good flat curves when sized to provide atleast 3 hours+ of burn time.

Lith-Ion do not and therefore to extract the useable stored engery one does need some form of voltage regulation as pointed out by charlie99.


Regards,




ch

 

[illumiGeek]

The 10 watt Solarc HID ballast is not regulated (but the higher power ones are).

At rated power, the 10W Solarc lamp has a median life of 1000 hours. Running it at 11W will increase output, but reduce lamp life by about 50%. Running at 8-9W will reduce output, but increase lamp life similarly. The "safe" power range for the 10 watt Solarc lamp is 8 watts to 11 watts. Anything more or less can significantly reduce lamp life.

As you can see, the 10 watt Solarc can benefit from regulation. But if you don't mind popping a new $80 lamp in your light several times a year, then it's not a problem.

A lot of people think that incandescent lamps can only be damaged by over-power, but even Tungsten Halogen/Xenon lamps can be damaged by low voltage operation. The "Halogen Scavenging Cycle" that redeposits lost Tungsten back on the filament needs very high temperatures to function. Running a Halogen lamp with low batteries can and will reduce both lamp life and output.

The big advantages of Li-Ion is long shelf life and good power to weight density. They are much more of a headache when it comes to charging and discharging limitations.

Aloha, iG

Oh, and BTW, unless something's changed, SLA batteries don't have a very flat discharge curve either...

 

[cool_hardware52]

The 10 watt Solarc HID ballast is not regulated (but the higher power ones are).

At rated power, the 10W Solarc lamp has a median life of 1000 hours. Running it at 11W will increase output, but reduce lamp life by about 50%. Running at 8-9W will reduce output, but increase lamp life similarly. The "safe" power range for the 10 watt Solarc lamp is 8 watts to 11 watts. Anything more or less can significantly reduce lamp life.

As you can see, the 10 watt Solarc can benefit from regulation. But if you don't mind popping a new $80 lamp in your light several times a year, then it's not a problem.

A lot of people think that incandescent lamps can only be damaged by over-power, but even Tungsten Halogen/Xenon lamps can be damaged by low voltage operation. The "Halogen Scavenging Cycle" that redeposits lost Tungsten back on the filament needs very high temperatures to function. Running a Halogen lamp with low batteries can and will reduce both lamp life and output.

The big advantages of Li-Ion is long shelf life and good power to weight density. They are much more of a headache when it comes to charging and discharging limitations.

Aloha, iG

Oh, and BTW, unless something's changed, SLA batteries don't have a very flat discharge curve either...

The voltage discharge curve for SLA batteries is not perfectly flat, but in capacities that permit 3-4 hour discharges the % of recoverable area under the curve is pretty high, not as flat as Ni-Cad though.

The Halide cycle in lamps does require high temps to work correctly, but the benefit of the Halide cycle is primarily the reduction in the deposition of evaporated tungsten on the inside of the bulb. The halide cycle does very little to increase the life of the filament. The tungsten that is redeposited on the filament is seldom replaced in the same areas that have the greatest evaporation rate. In other words the halide cycle keeps the lamp from darkening, but does not extend it's life. The effect of running halogen lamps below their design voltage is an extended filament life, and a darkened envelope.


Regards,



Tobin George

 

[Burner]

I still have some more testing to do, I can't tell if the voltage is dropping out, or if the charging circuit is limiting the charge because it has more batteries then it is expecting to have. I'm getting about 1/3 the time that I expect, and i have 2.5 times as many batteries as the circuit is expecting.